Jalview/JABAWS integration with Phytozome

Hi Jim, Peter, et.al.

Not exactly sure which list is most appropriate for this, so I’ll start with both.

Now that I’ve gotten past the hurtle of getting jabaws installed locally (and we’re moving forward with our DAS server to pass gene family data to Jalview), I wanted to resume discussions we first had back in 2011 regarding adding phylogenetic services to Jalview. At that time, Jabaws was in mid-transition from 1.x to 2.0, so I wanted to make sure my group re-engages on this a good picture of where things stand.

-silly question, but I wanted to verify that there hasn’t yet been a complete or partial integration of GBlocks, or any ML phylogenetic package (e.g, RAxML) into jabaws and jalview?

-is the thinking still that GBlocks should be a web service implementing the SequenceAnnotation interface?

-and the recommended way of adding a phylogenetic component is as a web service? Rather than being directly incorporated in the "Calculate Tree” set of NJ and Avg.Dist. tree builders in the Alignment Frame top menu? I imagine the answer is yes, if only to be able to execute cpu intensive phylo algorithms on bigger server-side hw, but I still wanted to check.

-is there additional documentation, besides the old threads with Joni in my group, for adding services to jabaws?

-you already have clients for tree rendering, with the ability to display bootstraps. I assume that would be client invoked to display results from a “tree build as web service” run? Or is there anything special about clients that handle web service output vs. others?

Also, just curious if there is someone currently on the Jalview project with exclusive or majority focus on Jabaws, now that Peter has moved on?

best,
-David

···


David M. Goodstein, Ph.D.
Phytozome Group Lead
Plant and Computational Genomics Group
Joint Genome Institute - U.S. Dept. of Energy
Center for Integrative Genomics - UC Berkeley

Hi David.

It’s great to hear that your system development plans have progressed.

To answer your last question first: Alexey Drozdetskiy has taken over the jabaws project, but is currently engaged in jpred development. We have no existing or partially implemented code for gblocks or phylogeny services, and so if you have developer time available, we (I) would be overjoyed if youd be able to work with us to implement these services.
The good news here is that Jalview’s core team will have tripled by the beginning of October, and so Mungo, Charles or myself will be able to create a client in Jalview for the new services.

In terms of a roadmap, I would create a client for new gblocks and also tcoffee’s alignment score services that annotates existing alignments, enabled as a preference in Jalview. Then, a new phylogeny service set in the web services menu would allow columns input to the analysis to be filtered according to these annotation. We can reuse the existing tree viewer for the results.

How does that sound ?

Jim.

···

On 24 Sep 2014 00:30, “David Goodstein” <dmgoodstein@lbl.gov> wrote:

Hi Jim, Peter, et.al.

Not exactly sure which list is most appropriate for this, so I’ll start with both.

Now that I’ve gotten past the hurtle of getting jabaws installed locally (and we’re moving forward with our DAS server to pass gene family data to Jalview), I wanted to resume discussions we first had back in 2011 regarding adding phylogenetic services to Jalview. At that time, Jabaws was in mid-transition from 1.x to 2.0, so I wanted to make sure my group re-engages on this a good picture of where things stand.

-silly question, but I wanted to verify that there hasn’t yet been a complete or partial integration of GBlocks, or any ML phylogenetic package (e.g, RAxML) into jabaws and jalview?

-is the thinking still that GBlocks should be a web service implementing the SequenceAnnotation interface?

-and the recommended way of adding a phylogenetic component is as a web service? Rather than being directly incorporated in the "Calculate Tree” set of NJ and Avg.Dist. tree builders in the Alignment Frame top menu? I imagine the answer is yes, if only to be able to execute cpu intensive phylo algorithms on bigger server-side hw, but I still wanted to check.

-is there additional documentation, besides the old threads with Joni in my group, for adding services to jabaws?

-you already have clients for tree rendering, with the ability to display bootstraps. I assume that would be client invoked to display results from a “tree build as web service” run? Or is there anything special about clients that handle web service output vs. others?

Also, just curious if there is someone currently on the Jalview project with exclusive or majority focus on Jabaws, now that Peter has moved on?

best,
-David


David M. Goodstein, Ph.D.
Phytozome Group Lead
Plant and Computational Genomics Group
Joint Genome Institute - U.S. Dept. of Energy
Center for Integrative Genomics - UC Berkeley


Jalview-discuss mailing list
Jalview-discuss@jalview.org
http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-discuss